Unsupervised machine learning detection of iceberg populations within sea ice from dual-polarisation SAR imagery

Evans, Ben ORCID: https://orcid.org/0000-0003-0643-526X; Faul, Anita ORCID: https://orcid.org/0000-0002-5911-2109; Fleming, Andrew ORCID: https://orcid.org/0000-0002-0143-4527; Vaughan, David G. ORCID: https://orcid.org/0000-0002-9065-0570; Hosking, J. Scott ORCID: https://orcid.org/0000-0002-3646-3504. 2023 Unsupervised machine learning detection of iceberg populations within sea ice from dual-polarisation SAR imagery. Remote Sensing of Environment, 297, 113780. 15, pp. https://doi.org/10.1016/j.rse.2023.113780

Before downloading, please read NORA policies.

Preview

Text (Open Access)
© 2023 The Author(s). Published by Elsevier Inc.
1-s2.0-S0034425723003310-main.pdf - Updated Version
Available under License Creative Commons Attribution 4.0.
Download (3MB) | Preview

Official URL: https://www.sciencedirect.com/science/article/pii/...

Abstract/Summary

Accurate quantification of iceberg populations is essential to inform estimates of Southern Ocean freshwater and heat balances as well as shipping hazards. The automated operational monitoring of icebergs remains challenging, largely due to a lack of generality in existing approaches. Previous efforts to map icebergs have often exploited synthetic aperture radar (SAR) data but the majority are designed for open water situations, require significant operator input, and are susceptible to the substantial spatial and temporal variability in backscatter that characterises SAR time-series. We propose an adaptive unsupervised classification procedure based on Sentinel 1 SAR data and a recursive Dirichlet Process implementation of Bayesian Gaussian Mixture Model. The approach is robust to inter-scene variability and can identify icebergs even within complex environments containing mixtures of open water, sea ice and icebergs of various sizes. For the study area in the Amundsen Sea Embayment, close to the calving front of Thwaites Glacier, our classifier achieved a mean pixel-wise F1 score against manual iceberg delineations from the SAR scenes of 0.960 ± 0.018 with a corresponding object-level F1 score of 0.729 ± 0.086. The method provides an excellent basis for estimation of total near-shore iceberg populations and has inherent potential for scalability that other approaches lack.

Item Type:	Publication - Article
Digital Object Identifier (DOI):	https://doi.org/10.1016/j.rse.2023.113780
ISSN:	00344257
Additional Keywords:	Iceberg, Machine learning, Automated, Bayesian, Classification, Radar
Date made live:	04 Sep 2023 08:43 +0 (UTC)
URI:	https://nora.nerc.ac.uk/id/eprint/535723

Item Type:

Publication - Article

Digital Object Identifier (DOI):

https://doi.org/10.1016/j.rse.2023.113780

ISSN:

00344257

Additional Keywords:

Iceberg, Machine learning, Automated, Bayesian, Classification, Radar

Date made live:

04 Sep 2023 08:43 +0 (UTC)

URI:

https://nora.nerc.ac.uk/id/eprint/535723